Download population dynamics of two species of kleptoparasitic spiders

2002. The Journal of Arachnology 30:31–38
POPULATION DYNAMICS OF TWO SPECIES OF
KLEPTOPARASITIC SPIDERS UNDER DIFFERENT
HOST AVAILABILITIES
Tadashi Miyashita: Laboratory of Biodiversity Science, School of Agriculture and
Life Sciences, University of Tokyo, Tokyo 113–8656 Japan.
Email: [email protected]
ABSTRACT. Kleptoparasitic spiders are known to have a close association with host spiders, yet there
have been few studies demonstrating how host availability influences the dynamics of kleptoparasites.
Field surveys were conducted at five sites differing in host composition in sub-tropical areas in Japan, at
about two-months intervals. Argyrodes flavescens and A. bonadea were both found more frequently on
webs of two Nephila species than expected from the web areas they occupied among webs of all web
spiders. Seasonal dynamics of Argyrodes changed greatly according to whether N. clavata was present or
not, indicating the importance of Nephila on Argyrodes populations. The peak density of A. bonadea came
earlier than that of A. flavescens. Because A. flavescens is known to limit the number of A. bonadea on
host webs, the decrease in the density of A. bonadea may be due to the effect of interspecific competition
by A. flavescens.
Keywords:
Interspecific competition, parasite, Theridiidae, phenology
A. flavescens O. Pickard-Cambridge 1880 and
A. bonadea (Karsch 1881), are commonly
found on orb-webs of various spider species.
Preliminary observations revealed that a large
number of the two species was found on the
webs of Nephila clavata and N. maculata (Fabricius 1793). These two Nephila species have
different phenologies, the former emerging
later in the season. Also, the distributions of
the two species differ, N. maculata lives all
over the island while N. clavata inhabits only
the northern part of the island. These characteristics enable us to estimate the importance
of host availability.
In this paper I examined seasonal dynamics
of the two Argyrodes species and their potential host spiders in five sites differing in host
species compositions. I hypothesize that 1) the
Nephila species are the preferred host for Argyrodes, and 2) the presence or absence of
Nephila clavata changes population dynamics
of Argyrodes greatly.
Spiders in the genus Argyrodes Simon 1864
are often called kleptoparasites that rely almost exclusively on host spiders. However,
the interaction between Argyrodes and their
hosts is complex: Argyrodes can be kleptoparasites, commensals, or predators to their
hosts depending on the species and host conditions (e.g., Trail 1980; Tanaka 1984; Larcher
& Wise 1985; Whitehouse 1986; Cangialosi
1997). These close associations lead to the inference that population dynamics of Argyrodes is highly dependent on host availability.
Vollrath (1987) demonstrated the dynamics of
two Argyrodes species and their host Nephila
clavipes (Linnaeus 1767). However, there
have been no studies showing how changes in
host species composition influences Argyrodes dynamics. Ideally such a causal relationship should be evaluated by field experiments
in which a particular host is removed, but population manipulation on a large scale sufficient to change host availability is impractical.
Examining the variation of kleptoparasite dynamics in areas differing in host availability
seems to be the only way to estimate the
strength of associations between the host and
kleptoparasite at the population level.
On Okinawa Island located in the southwestern part of Japan, two Argyrodes species,
METHODS
Field studies were conducted at four sites
on Okinawa Island, Nakagusuku (NK), Nangusuku (NG), Shoshi (SH), Nakijingusuku
(NJ), and at one site on Iheya Island (IH),
southwestern Japan. The climate of these sites
31
32
THE JOURNAL OF ARACHNOLOGY
is sub-tropical, with an average temperature of
22.6 8C and an annual rainfall of 2,100 mm.
Previous observations revealed that Nephila
clavata, a major host for Argyrodes, lives at
all sites except for NK which is located at the
southern part of Okinawa Island. Iheya Island
lies about 60 km away from Okinawa Island
and is much smaller (about 20 km2) than that
of Okinawa Island (about 1,199 km2). Thus I
expected densities and/or species richness of
host spiders to be low on that island.
I established two 50 m lines along the roadside in forests at each site, and surveyed these
transects five times (four times for IH) from
July 1997 to April 1998 at intervals of approximately 2–3 mo. I recorded all web spiders with body lengths larger than 4 mm living within 2 m from the transect and 2 m from
the ground. I recorded the spider body length,
the vertical and horizontal web diameters, any
prey being consumed, and any small insects
remaining in the web. At the same time I
counted and recorded the body length of all
Argyrodes found on the spider webs with calipers. Because it was difficult to know the
stage of Argyrodes in the field, I used body
length as an indicator for the stage and determined the smallest size of adults, viz., 3 mm
for A. flavescens and 2.5 mm for A. bonadea.
In this study, web areas of hosts instead of
density of host individuals was used for representing host availability because the body
size of host spiders varies greatly depending
on the species and seasons. For instance, the
web of Nephila maculata is much larger than
that of most other spiders, so that the amount
of habitat and food resources available to Argyrodes may vary greatly.
The density of each Argyrodes on a particular host species (no./cm2) was calculated as
follows, S Ni /S Ai, where Ni is the number of
Argyrodes on the ith web of a particular host
species and Ai is the web area of the ith web.
Data from the two transects were combined at
each site. A binomial test was performed to
ascertain whether each Argyrodes prefers a
particular host spider species. The null hypothesis was that the total number of Argyrodes on webs of host species j (Nj) is determined by the proportion of the web area of
that host, i.e.,
Nj 5 Aj @
O A · O N,
j
j
where Aj is the total web area of host species
j and Nj is the total number of Argyrodes sp.
on host species j. Hosts harboring at least ten
Argyrodes or those occupying more than 30%
of the total web areas of all species combined
were used for the analysis. Significance level
was adjusted by the sequential Bonferroni
method (Sokal and Rohlf 1995) for each season.
RESULTS
Dynamics of Argyrodes and their host.—
The kleptoparasitic spiders found in this survey were mostly A. flavescens and A. bonadea. Figure 1 shows seasonal changes in the
density of the two Argyrodes spp. and web
areas of host spiders in five study sites. In NK
where N. clavata is absent, the dynamics of
web area were mostly determined by N. maculata. Although the peak density of A. flavescens was delayed relative to that of host
web area, its density decreased abruptly in
November when N. maculata disappeared. In
NG, SH and NK, however, N. clavata is abundant in November, and Argyrodes maintained
high densities in November. It is noteworthy
that the timing of peak density of the two Argyrodes species differed, with A. bonadea
reaching the peak earlier. Another important
point is that both Argyrodes densities declined
to low levels in January and April when web
areas of spiders other than Nephila remained
at levels similar to those in November. The
main species in this period were Leucauge
blanda (L. Koch 1878), Cyclosa confusa Bosenberg & Strand 1906 and Neoscona scylla
(Karsch 1879). In IH where N. clavata was
not found on the census route, the density of
Argyrodes showed a peak in August and then
dropped sharply in November, which is similar to the situation in NK. Unlike other study
sites, the density of A. bonadea was much
higher than that of A. flavescens.
Percentage of adult Argyrodes.—The percentage of adult individuals was calculated
separately for areas where both Nephila species were present (NG, SH, NJ) and Nephila
clavata was absent (NK, IH). In areas where
both Nephila species were present, seasonal
change in the percentage of adults differed
greatly between the two Argyrodes (Fig. 2).
The peak occurrence of adult A. flavescens
was found in autumn while that of A. bonadea
was in spring and early summer. In areas
where Nephila clavata was absent, the differ-
Figure 1.—Seasonal dynamics of the numbers of the two Argyrodes species, and web areas of the two Nephila species and other web spiders in 5 sites in
Okinawa.
MIYASHITA—DYNAMICS OF KLEPTOPARASITIC SPIDERS
33
34
THE JOURNAL OF ARACHNOLOGY
Figure 2.—Percentage of adult Argyrodes in areas with and without Nephila clavata. Numerals in the
figure represent sample size.
ence was less conspicuous due to the high percentage of adult A. flavescens in spring. Argyrodes bonadea showed a similar seasonal
pattern in the two types of areas.
Host preference of Argyrodes.—A total of
18 species of web spiders was recorded including all sites and seasons (Table 1). Twelve
species were found to be hosts for A. flavescens, and 14 were hosts for A. bonadea.
The results of the binomial tests (Table 2)
revealed that A. flavescens had a strong tendency to prefer Nephila clavata in August and
November, although no preference was found
in July when Nephila clavata was still small
in body size. Nephila maculata was preferred
in July probably because Nephila was already
large. Besides the two Nephila species, only
Argiope minuta was preferred significantly.
Leucauge blanda and Cyclosa confusa had
significantly fewer A. flavescens on their webs
than expected from their web areas.
Argyrodes bonadea also preferred N. clavata in August and November, but the preference for N. maculata was obscure. Unlike
A. flavescens, Cyrtophora moluccensis, which
builds large, horizontal and tightly-meshed
webs, was a preferred host for A. bonadea.
Leucauge blanda had fewer A. bonadea than
expected from its web area.
Prey density on host webs.—Figure 3
shows seasonal changes in prey density on
Nephila webs. In all areas, prey density on
host webs increased in November and then decreased markedly in January. The main prey
in November were small plant hoppers in all
study sites.
DISCUSSION
The dynamics of the two Argyrodes species
were largely affected by the phenologies of
Nephila species. This is clearly illustrated by
the following findings: 1) where there is no
N. clavata (NK and IH), densities of Argyrodes declined rapidly in November when N.
maculata had disappeared; but where both Nephila species were present, the density increased in November for A. flavescens and decreased slightly for A. bonadea; 2) preference
for Nephila webs was prominent, especially
by A. flavescens. The first finding is particularly good evidence for the importance of Nephila on Argyrodes dynamics because this can
be viewed as a ‘‘natural experiment’’ in which
one major host was ‘‘removed’’. Grostal &
Walter (1999) revealed a similar result showing that Nephila spp. were the preferred hosts
for A. antipodianus. The reason why Nephila
webs are suitable for Argyrodes may be that
these webs have a barrier web on both sides
of the capture web, providing a space for living. The significance of barrier webbing for
Argyrodes has already been argued by several
researchers (Whitehouse 1986 & 1997; Cangialosi 1997). Comparing the two Nephila
species, N. clavata has a more elaborate barrier web and N. maculata sometimes lacks a
barrier web (unpub. obs.). This may explain
why N. clavata was preferred more frequently
2
2
Ar
1
2
1
2
2
2
1
2
Cm
1
2
Cu
1
2
1
1
2
1
2
1
1
2
Gm
1
2
1
1
1
1
1
1
1
2
Lb
2
1
1
1
1
1
1
1
2
2
Ns
1
2
2
1
2
1
1
2
2
2
Nsu
Host species
1
1
Nt
2
2
Er
1
1
1
2
1
1
1
1
1
1
Cc
1
1
1
2
2
2
Cm
2
1
2
1
Aa
1
1
2
1
1
2
Am
2
2
Te
2
2
2
2
Ac
Nm 5 Nephila maculata, Ar 5 Araneus sp., Gm 5 Gasteracantha mammosa, Nsu 5 Neoscona subpullata, Cc 5 Cyclosa confusa, Am 5 Argiope minuta, Nc
5 Nephila clavata, Cm 5 Cyrtophora moluccensis, Lb 5 Leucauge blanda, Nt 5 Neoscona theisi, Cm 5 Cyclosa mulmeinensis, Te 5 Tetragnatha sp., Av
5 Araneus ventricosus, Cu 5 Cyrtophora unicolor, Nc 5 Neoscona scylla, Er 5 Eriophora yanbaruensis, Aa 5 Argiope aemula, and Ac 5 Achaearanea sp.
1
1
2
1
Av
Argyrodes bonadea
1
NK
1
NG
1
1
SH
1
1
NJ
1
1
IH
Nc
1
1
2
2
Nm
Argyrodes flavescens
1
NK
1
NG
1
1
SH
1
2
NJ
1
1
IH
Site
Table 1.—Host spiders for the two Argyrodes species found in the census route in each study site (all seasons pooled). (1) 5 Argyrodes found, (2) 5
Argyrodes not found, blank 5 host not found.
MIYASHITA—DYNAMICS OF KLEPTOPARASITIC SPIDERS
35
36
THE JOURNAL OF ARACHNOLOGY
Table 2.—Results of binomial test to assess whether each Argyrodes species prefers a particular host
spider. The null hypothesis is that an Argyrodes chooses a host spider in proportion to its web area.
Positive signs mean preference (111 5 p , 0.001, 11 5 p , 0.01, 1 5 p , 0.05), negative signs
mean avoidance (222 5 p , 0.001, 22 5 p , 0.01, 2 5 p , 0.05).
Abbreviations for species name are the same as in Table 1.
Argyrodes flavescens
Host species
Site
Month
Nm
NK
NG
Aug
Jul
Aug
Nov
Jul
Aug
Nov
Jul
Aug
Nov
Aug
111
ns
SH
NJ
IH
111
ns
ns
Nc
Av
Cm
Lb
Ns
Am
2
ns
ns
111
++
ns
111
111
Cc
ns
222
ns
ns
ns
ns
ns
11
222
ns
111
ns
1
Cc
Am
2
Argyrodes bonadea
Host species
Site
Month
Nm
NK
NG
Aug
Jul
Aug
Nov
Jul
Aug
Nov
Jul
Aug
Nov
Aug
11
ns
ns
SH
NJ
IH
ns
222
ns
Nc
Av
Cm
Lb
Ns
2
ns
111
1
ns
11
ns
ns
111
1
ns
ns
22
ns
ns
ns
ns
2
ns
ns
ns
than N. maculata by the two Argyrodes. Other
studies have shown that Argyrodes spiders often live in Nephila webs (e.g., Elgar 1993) but
only a few of them have demonstrated the seasonal dynamics of Nephila-Argyrodes systems. Vollrath (1987) showed that the dynamics of A. elevatus and A. caudatus were
closely related to the number of N. clavipes in
Panama, yet no evidence was provided showing that alternative host spiders were not responsible for this pattern. The present study
clearly demonstrates the importance of Nephila as hosts for Argyrodes both from the
‘‘natural experiment’’ and preference analysis.
Another interesting point in the population
dynamics of the two Argyrodes is that the
peak density of A. bonadea was earlier than
ns
1
1
1
ns
ns
ns
ns
that of A. flavescens. This difference may be
due to interspecific competition between the
two Argyrodes species. Miyashita (2001)
demonstrated that experimental removal of A.
flavescens increased number of A. bonadea
within only 2 d, and the number of individuals
removed was positively correlated with the
rate of increase in A. bonadea. Although this
experiment indicated interspecific competition
on a small spatial scale, this effect is likely to
extend up to regional scales. There is additional circumstantial evidence supporting the
likelihood of interspecific competition. In a
temperate area where only A. bonadea was
present, Yasuda (pers. comm.) found that this
species is most abundant in mid-October
which is the late adult period of N. clavata
MIYASHITA—DYNAMICS OF KLEPTOPARASITIC SPIDERS
Figure 3.—Seasonal change in prey density (no./
web area) on host spider webs. The 3 sites shown
here have both of the two Nephila species.
and 2 mo behind the peak of the web area.
On Okinawa, however, A. bonadea showed a
peak in late August and the density decreased
in November, the beginning of the adult period in Okinawa. This means that A. bonadea
living in regions without A. flavescens showed
host-parasite dynamics similar to A. flavescens. Accordingly, interspecific competition
by A. flavescens may have shifted the dynamics of A. bonadea to earlier seasons. An alternative explanation is that the life cycle of A.
bonadea in sub-tropical Okinawa is different
from that of the temperate area, which is unrelated to competitive interactions with A. flavescens. It would be difficult to test whether
there is a causal relationship between the dynamics of the two Argyrodes species because
large scale experimental manipulation is necessary.
Host composition appears to have a minor
impact on the age structure of Argyrodes, especially for A. bonadea (Fig. 2). The increase
in the proportion of adults in spring in areas
without N. clavata might have been an artifact
due to the small sample size (N 5 8). Adults
of A. flavescens were found throughout the
year but the peak was in autumn when prey
on host webs were most abundant. If the increase in prey availability had already been
initiated in October, this can explain the higher proportion of adults in November, because
spiderlings of A. flavescens can reach maturity
and produce egg sacs in a month (unpub.
obs.). Contrarily, adults of A. bonadea were
rarely found in autumn, suggesting A. bonadea does not accerelate its developmental rate
37
by consuming prey abundant on host webs but
adjusts its life cycle by retarding its development. It appears that A. flavescens has a more
opportunistic life cycle strategy than A. bonadea.
From winter to spring when the main hosts
were absent, Argyrodes spiders were found to
live on the webs of smaller orb-weavers such
as Leucauge blanda, Cyclosa confusa and
Gasteracantha mammosa, which were not
preferred in the presence of Nephila. Whitehouse (1988) found that the host range of A.
antipodiana in winter was larger than that in
growing and reproductive seasons, probably
because overwintering individuals need webs
for shelter only. However, the situation of Argyrodes in the present study is quite different
because I often observed Argyrodes ingesting
silk of host webs as well as stealing prey (unpub. obs.). The difference is mainly due to
climate conditions: the present study was
made in the sub-tropics whereas the study by
Whitehouse was in a temperate region. Although it seems unlikely that sub-tropical Argyrodes can obtain sufficient food in winter,
they may be able to survive the harsh period
by living on unfavorable hosts. Grostal &
Walter (1999) found no A. antipodianus on
webs of Leucauge sp. and Gasteracantha sp.
in Australia but the season they studied was
the period when the main host Nephila was
present. It seems necessary to clarify the role
of non-preferred host for the maintenance of
Argyrodes populations.
ACKNOWLEDGEMENTS
I thank Mary Whitehouse for comments on
the manuscript, and Akira Shinkai, Takafumi
Chida, Aya Shimazaki, Yasunori Maezono
and Claire Cartan for assistance in the field.
This research was supported by Fujiwara Natural History Foundation.
LITERATURE CITED
Cangialosi, K.R. 1997. Foraging versatility and the
influence of host availability in Argyrodes trigonum (Araneae, Theridiidae). Journal of Arachnology 25:182–193.
Elgar, M.A. 1993. Inter-specific associations involving spiders: Kleptoparasitism, mimicry and mutualism. Memoirs of the Queensland Museum
33:411–430.
Grostal, P. & D.E. Walter. 1999. Host specificity
and distribution of the kleptoparasitic spider Argyrodes antipodianus (Araneae, Theridiidae) on
38
orb webs in Queensland, Australia. Journal of
Arachnology 27:522–530.
Larcher, S.R. & D.H. Wise. 1985. Experimental
studies of the interactions between a web-invading spider and two host species. Journal of Arachnology 13:43–59.
Miyashita, T. 2001. Competition for a limited space
in kleptoparasitic Argyrodes spiders revealed by
field experiments. Population Ecology 43:97–
103.
Sokal, R.R. & F.J. Rohlf. 1995. Biometry, 3rd ed.,
W.H. Freeman, New York.
Tanaka, K. 1984. Rate of predation by a kleptoparasitic spider, Argyrodes fissifrons upon a large
host spider, Agelena limbata. Journal of Arachnology 12:363–367.
Trail, D.S.S. 1980. Predation by Argyrodes (Theridiidae) on solitary and communal spiders. Psyche 87:349–355.
THE JOURNAL OF ARACHNOLOGY
Vollrath, F. 1987. Kleptobiosis in spiders. In Ecophysiology of Spiders. Nentwig W. ed, pp 274–
286, Springer-Verlag, Berlin.
Whitehouse, M.E.A. 1986. The foraging behaviours
of Argyrodes antipodiana (Theridiidae), a kleptoparasitic spider from New Zealand. New Zealand Journal of Zoology 13:151–168
Whitehouse, M.E.A. 1988. Factors influencing
specificity and choice of host in Argyrodes antipodiana (Theridiidae, Araneae). Journal of Arachnology 16:349–355.
Whitehouse, M.E.A. 1997. The benefits of stealing
from a predator: foraging rates, predation risk,
and intraspecific aggression in the kleptoparasitic
spider Argyrodes antipodiana. Behavioural Ecology 8:663–667.
Manuscript received 1 August 2000, revised 10
May 2001.